{Lojban(1990), tr. Bob LeChevalier (`lojbab')}

loi bangu cu litki\\
.inaja ri sutfle fi ti\\
.iku'i na go'i\\
.ili'i nunsma semau ro valsi\\
temau leka zanselsku

.i loi valsi cu duksligu\\
.i ri na sutra co banzu\\
le mu'e kavbu le besysutra\\
poi sutfau\\
gi'e ba purci

.i mi djica lenu penmi do\\
ca noda vi node\\
ma'inai rodi\\
ba'o ro jagdimna

.i .e'u mi'o xruti fi le dinju\\
pe vi la cmalu ke stici gaimoi\\
.i na'e darno\\
.i le rirxe cu zvati\\
.i le solri .e le vanbi\\
cu no'e se zgana\\
.i mi'o vu zutse va'o \\
lei smaji poi sutflefau\\
gi'e ba purci

.i .ai mi banoroi pilno loi valsi\\
.i ri na smuni lemi selsmu\\
gi'e na velsku lemi selsku\\
gi'e pilka le smuni sekai \\
le baltutra nenri\\
poi noroi se pencu\\
gi'e noroi se jicla\\
gi'e noroi mecrai se pagre
First some notes:

\item{1.} My translation is not quite as literal as Nick's appears to
be (not being
familiar with the other three ALs).  I have tried to maintain a sense of the
style, denotation, and connotation, of the words used.  However, Lojban is
{\it not\/} an Indo-European language, and certain things must be
rephrased in order
to be both (unambiguously) grammatical and to capture the meaning correctly.
\item{2.} Lojban is less tolerant of metaphor than other languages,
but does allow
analytic metaphors (where the predicate place structures are semantically
preserved in the combination).
\item{3.} Nick describes the text as a song.  I saw no apparent match in rhythm
and/or syllable count between lines of the English and the AL versions.  I
presume therefore that the translation is in free verse and is not intended
to match the music (which I don't know anyway).
\item{4.} Pronunciation: Romance (Esperanto) vowels, except {\tt y},
which is `schwa'
(neutral medial never-stressed).  Vowel pairs without an apostrophe between
them are diphthongs which analytically break up into the two vowels.  The
apostrophe indicates that the two vowels are in separate syllables separated
by a {\sl devoiced glide}.  English speakers can do fine by pronouncing the
apostrophe as {\sf /h/}.  Most consonants are IPA; {\tt r} can range
from trill to
flap.  {\tt c} is pronounced as English {\sf /sh/}; {\tt j} is the
voiced equivalent {\sf /zh/},
as the medial consonant of Engliah `measure'.  Thus {\tt dj} is
English {\sf /j/}.
{\tt x} is a velar fricative (greek $\chi$), preferably unvoiced.  Stress on
predicate words (the longer ones) is penultimate ignoring any {\tt y}
Period indicates a mandatory pause, which can be as short as a glottal stop
(which is an allophone of pause).  The title is thus pronounced:

/meh,LAH,heh  loo,BAHN,goo,lee,hoo  nee,ho/

Following is a literal English translation of the Lojban:

{\parskip=4pt plus 1pt minus 1pt
That represented by ``Language''. New topic.

(The mass of) Language is liquid,\\
only if it (language) fast-flows to here.\\
But not-true, the latter.\\
Abstract-experiencing-of event-of-silence which-is-more-than each (any) word,\\
more-in the property of ameliorative(good)-being-expressed.

(The mass of) Words are excess-solid\\
They (words) are not quick such-that sufficient\\
in the abstract-achievement of capturing the brain-[quick-thing]\\
which quickly-occurs\\
and then is-past.

I desire the event of meeting you\\
during no-when, at nowhere,\\
according-to-reference-frame none\\
inthe aftermath of all result-dooms.

(Suggestion!) We return (ourselves-elliptical) to the building\\
which is at that called 'Little type-of West Twelfth-thing'.\\
Other-than far (it is - elliptical).\\
The river is at (it-elliptical).\\
The sun and the environment\\
are neutrally-other-than observed (neither extreme of observed/non-observed)\\
We there-yonder sit in-environment\\
this (mass of) silence which swift-flowingly-happens\\
and then is past.

(Intention!) I in-the-future-never use (of the mass of) Words.\\
They (words) do-not mean my thing-meant\\
and are-not forms-of-expressing my things-expressed\\
and are skins of the meaning, characterized by\\
the grand-territory inside\\
which never is touched\\
and never is stirred\\
and never least-superlative passed-through.

Now a note/complain/what have you which I think is most revealing of the
nature and 'neutrality' of the other three languages.  In the translations
of the last verse, specifically:

\item{\quad}``I won't use words again\\
they don't mean what I mean\\
they don't say what I say ...''

Nick has translated both occurances of English words `mean' and `say' with
the same counterpart in each of the other three languages.  But the English
words do not denote the same thing.  When a person says something, it is
different from when words say something.  Likewise, the `meaning' of words
is semantically distinct differ from the `intended meaning' of one who might
use the words.  This is intuitive to an English speaker, who knows the range
of meaning of the words.

If each of the other ALs use the same word to capture both senses of `mean'
and `say', then I assert that they are flawed and biased towards English
and/or every other language that blurs this distinction.  I suspect that
such blurring, if in other languages, will tend to be only in the culturally
similar European ones (Opinions/confirmation from those expert in non-
European languages, please?)  If Esperanto, Ido, and Volap\"uk all borrowed
European roots along with their complete semantic baggage, then those
languages are going to be inherently less understandable to a non-European
who does not share the cultural background.  This is a particularly
insidious kind of bias because, as one Esperantist has pointed out, it seems
that both the European and non-European are having `the shared experience'
of acquiring the AL they both learn.  But for one learner, it is
predominantly a regularized, simplified, form of their own language; for the
other, the subtle semantics needed for poetry is not shared.  (This
criticism applies more obviously for BASIC English, since people can easily
see the confusing semantic range of the word-plus-preposition combinations
that make that language work.)

I do not claim that the difficulty is insurmountable.  Certainly non-
Europeans have written poetry in Esperanto that was understood and
appreciated by Europeans, possibly in a way that is not as easily possible
if the European had to learn the native language of the poet, which has a
much heavier cultural/connotative load.  I suspect that (European) speakers
who can converse in Esperanto fluently with {\it non-Europeans}, and who
therefore are thinking in Esperanto rather than translating from their
native language as they go, have largely bypassed this difficulty.

I also accept, not as a flame to him, that Nick may not have translated the
poem incorrectly, or at least not ideally, in that the ALs may indeed have
separate words for each of the distinct meanings in the song.  But most
Esperantists and other ALers are NOT going to know the acquired language
thoroughly and will often transfer native language semantics to the highly
cognate forms of the AL.  (This problem occurs just as severely in acquiring
`national' languages.)

By the way, the words is this poem are not the issue.  Countless other
examples can be found.  Do the Esperanto color-words correspond to one or
another `national language'?  How does it distinguish between the two
Russian `blues' or the red/pink distinction of English?  (I use Esperanto
because I suspect that the question is answerable by readers, not to single
it out for criticism among the others.)  Does the Esperanto word for `arm'
include the hand, as in Russian, or exclude it as in many other languages?
Is a Russian taught Esperanto informed of this possible confusion in talking
with non- Russians.  The word `morning' is even more extreme, since the
hours implied by the word for the concept in each language vary
considerably.  (A test!  Is it legit for an Esperantist to say the
equivalent of 1-in-the-morning for 1AM, and 11-in-the-morning for 11AM?  Is
an English student of Esperanto told this when taught the word for morning
(or the method of telling time.)

My points can be summarized as two:
\item{1.} I agree with whoever it was that criticized ALs implicitly as being
languages that people think they know after finishing the textbook.
\item{2.} As a corollary, it is a disadvantage for an AL to be `much
like' any other
single language in particular.  The speakers of that language have either a
benefit or a handicap, depending on how you look at it; they have an easier
time learning subtle features of the AL and a harder time recognizing the
differences that {\it must\/} be present for it to be an effective
communications tool.  The former is an unfair bias; the latter calls into
question whether the AL is suitable as an IL.

Since I lead the Lojban effort, I of course (biasedly?) support Lojban as
overcoming these issues.  Lojban is just as easy to learn as other ALs with
lots of regularity and simplification.  But since the language is tied to a
predicate grammar strikingly different from any other language, a speaker
translating anything but the simplest statements must significantly
reformulate the expression (as shown in the translation above) in order to
properly express it in Lojban.  The result is easily understood to another
Lojban speaker, and indeed in back-translation, is not too difficult in
English.  But a Lojbanist {\it must\/} think clearly about what s/he is saying in
order to even say the sentence; those who use other ALs do not necessarily
do so.  Thus, I think Lojban aids a learner in acquiring the `different
perspective' of a second language, and a Lojbanist who gets by the initial
hurdle of unfamiliar words and structures more rapidly acquires that added
competence that is considered `knowing' a language.

I am interested in what professional linguists have to say about translation
of semantics of this sort, and also how those working in machine translation
and comp. ling. deal with what must be an especially troubling problem.